
# connect to GDrive
from google.colab import drive
drive.mount('/content/drive', force_remount=True)

Mounted at /content/drive


#
# Initialize Basic File paths...
#

import os
root_dir_fp    = "/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/"
data_fp        = os.path.join(root_dir_fp, "data")
model_fp       = os.path.join(root_dir_fp, "models")
output_fp      = os.path.join(root_dir_fp, "output")

#
# Initialize Configuration File paths...
#
model_save_fp         = model_fp + '/pokermate_stage2_objectDetection.h5'


# Run the following to install required packages
# !pip install pandas numpy tensorflow keras scikit-learn github gitpython pylabel ultralytics


!pip install ultralytics omegaconf

Collecting ultralytics
  Downloading ultralytics-8.0.222-py3-none-any.whl (653 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 654.0/654.0 kB 4.1 MB/s eta 0:00:00
Collecting omegaconf
  Downloading omegaconf-2.3.0-py3-none-any.whl (79 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 79.5/79.5 kB 11.0 MB/s eta 0:00:00
Requirement already satisfied: matplotlib>=3.3.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (3.7.1)
Requirement already satisfied: numpy>=1.22.2 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (1.23.5)
Requirement already satisfied: opencv-python>=4.6.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (4.8.0.76)
Requirement already satisfied: pillow>=7.1.2 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (9.4.0)
Requirement already satisfied: pyyaml>=5.3.1 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (6.0.1)
Requirement already satisfied: requests>=2.23.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (2.31.0)
Requirement already satisfied: scipy>=1.4.1 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (1.11.4)
Requirement already satisfied: torch>=1.8.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (2.1.0+cu118)
Requirement already satisfied: torchvision>=0.9.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (0.16.0+cu118)
Requirement already satisfied: tqdm>=4.64.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (4.66.1)
Requirement already satisfied: pandas>=1.1.4 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (1.5.3)
Requirement already satisfied: seaborn>=0.11.0 in /usr/local/lib/python3.10/dist-packages (from ultralytics) (0.12.2)
Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from ultralytics) (5.9.5)
Requirement already satisfied: py-cpuinfo in /usr/local/lib/python3.10/dist-packages (from ultralytics) (9.0.0)
Collecting thop>=0.1.1 (from ultralytics)
  Downloading thop-0.1.1.post2209072238-py3-none-any.whl (15 kB)
Collecting antlr4-python3-runtime==4.9.* (from omegaconf)
  Downloading antlr4-python3-runtime-4.9.3.tar.gz (117 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 117.0/117.0 kB 16.4 MB/s eta 0:00:00
  Preparing metadata (setup.py) ... done
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (1.2.0)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (0.12.1)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (4.45.1)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (1.4.5)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (23.2)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (3.1.1)
Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.3.0->ultralytics) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.1.4->ultralytics) (2023.3.post1)
Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests>=2.23.0->ultralytics) (3.3.2)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests>=2.23.0->ultralytics) (3.6)
Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests>=2.23.0->ultralytics) (2.0.7)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests>=2.23.0->ultralytics) (2023.11.17)
Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (3.13.1)
Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (4.5.0)
Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (1.12)
Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (3.2.1)
Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (3.1.2)
Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (2023.6.0)
Requirement already satisfied: triton==2.1.0 in /usr/local/lib/python3.10/dist-packages (from torch>=1.8.0->ultralytics) (2.1.0)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7->matplotlib>=3.3.0->ultralytics) (1.16.0)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch>=1.8.0->ultralytics) (2.1.3)
Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch>=1.8.0->ultralytics) (1.3.0)
Building wheels for collected packages: antlr4-python3-runtime
  Building wheel for antlr4-python3-runtime (setup.py) ... done
  Created wheel for antlr4-python3-runtime: filename=antlr4_python3_runtime-4.9.3-py3-none-any.whl size=144554 sha256=f66916cc45cbb0cca88d363f35533d82cf260de368e027a7e3a5932af0fe9fed
  Stored in directory: /root/.cache/pip/wheels/12/93/dd/1f6a127edc45659556564c5730f6d4e300888f4bca2d4c5a88
Successfully built antlr4-python3-runtime
Installing collected packages: antlr4-python3-runtime, omegaconf, thop, ultralytics
Successfully installed antlr4-python3-runtime-4.9.3 omegaconf-2.3.0 thop-0.1.1.post2209072238 ultralytics-8.0.222


import os
import sys
import ntpath
import shutil
import pandas as pd
import numpy as np
from PIL import Image, ImageDraw
from IPython.display import Image as ImageDisplay
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
import tensorflow as tf
import keras
import torch
import cv2
from pathlib import Path
from ultralytics import YOLO

%matplotlib inline


## Setting up the dataset

images       = []
annotations  = []
data_yolo    = data_fp + "/stage2_images/"
train_fp     = data_fp + "/stage2_images/train/"
test_fp      = data_fp + "/stage2_images/test/"
val_fp       = data_fp + "/stage2_images/val/"


def setup_dataset_dir(target_fp):
  for file in os.listdir(target_fp):
    file_ext = ntpath.basename(file).rsplit(".", 1)[1].strip().lower()
    if file_ext in ['jpg', 'jpeg', 'png', 'webmp']:
      # image
      images.append(os.path.join(target_fp, file))
    elif file_ext in ['xml', 'exml', 'json']:
      # annotation
      annotations.append(os.path.join(target_fp, file))
    else:
      pass
    # copy images and annotations to a new dir setup
    dest_images_fp = os.path.join(target_fp, "images")
    dest_labels_fp = os.path.join(target_fp, "labels")
    os.mkdir(dest_images_fp)
    os.mkdir(dest_labels_fp)
    for img_fp in images:
      print("moving: {}".format(ntpath.basename(img_fp)))
      shutil.move(img_fp, dest_images_fp)
    for ann_fp in annotations:
      print("moving: {}".format(ntpath.basename(ann_fp)))
      shutil.move(ann_fp, dest_labels_fp)
    return f"Done: {target_fp}!"

def rename_files(images_fp, labels_fp):
  failed = []
  i=0
  for img in tqdm.tqdm(os.listdir(images_fp)):
      file_name = ntpath.basename(img).rsplit(".", 1)[0]
      img_fp = images_fp + img
      label_fp = labels_fp + file_name + ".txt"
      try:
          if Path(img_fp).exists() and Path(label_fp).exists():
              os.rename(img_fp, f"{images_fp}image_{i}.jpg")
              os.rename(label_fp, f"{labels_fp}image_{i}.txt")
          else:
              print("Failed - ", i, img)
              failed.append(img_fp)
      except Exception as e:
          print("ERROR::::", e, i, img)
      i+=1


# Original dataset had arbitary filenames, cleaned and renamed all the files
# reduced the size of the dataset.
# fp = "./train"
# images_fp = fp + "/images/"
# labels_fp = fp + "/labels/"
# rename_files(images_fp, labels_fp)

# fp = "./val"
# images_fp = fp + "/images/"
# labels_fp = fp + "/labels/"
# rename_files(images_fp, labels_fp)

# fp = "./test"
# images_fp = fp + "/images/"
# labels_fp = fp + "/labels/"
# rename_files(images_fp, labels_fp)


# Convert xml annotations to YOLO txt version (label r, t, l, b coords)
#from pylabel import importer
#dataset = importer.ImportVOC(path=train_fp + "labels/")
#dataset.export.ExportToYoloV5(output_path=train_fp + "labels")
#dataset = importer.ImportVOC(path=test_fp + "labels/")
#dataset.export.ExportToYoloV5(output_path=test_fp + "labels/")


## Display

train_data  = data_yolo + "train/labels"
test_data   = data_yolo + "test/labels"
val_data    = data_yolo + "val/labels"

print("::Total Images::\n\n")
print(f"Training Data:\t\t{len(os.listdir(train_data))}")
print(f"Testing Data:\t\t{len(os.listdir(test_data))}")
print(f"Validation Data:\t{len(os.listdir(val_data))}")

::Total Images::


Training Data:		4000
Testing Data:		506
Validation Data:	700


# Function to convert YOLOv5 format coordinates to OpenCV format
def yolo_to_opencv(box, img_shape):
    x, y, w, h = box
    x_center, y_center = x * img_shape[1], y * img_shape[0]
    box_width, box_height = w * img_shape[1], h * img_shape[0]
    x_min, y_min = int(x_center - box_width / 2), int(y_center - box_height / 2)
    x_max, y_max = int(x_center + box_width / 2), int(y_center + box_height / 2)
    return x_min, y_min, x_max, y_max

def visualize_image(image_fp, label_fp):
  # Path to label and image files
  label_path = label_fp
  img_fp = image_fp

  # Read image
  img_arr = cv2.imread(img_fp)
  img_shape = img_arr.shape[:2]

  # Display the original image
  plt.subplot(1, 2, 1)
  plt.imshow(cv2.cvtColor(img_arr, cv2.COLOR_BGR2RGB))
  plt.title("Original Image")
  plt.axis('off')

  # Read YOLOv5 format label file
  with open(label_path, 'r') as file:
      lines = file.readlines()

  # Iterate through each line in the label file
  for line in lines:
      # Split the line into label, x, y, w, h
      label, x, y, w, h = map(float, line.strip().split())

      # Convert YOLOv5 format to OpenCV format
      x_min, y_min, x_max, y_max = yolo_to_opencv((x, y, w, h), img_shape)

      # Draw bounding box on the image
      color = (0, 255, 0)  # Green color
      thickness = 2
      img_arr = cv2.rectangle(img_arr, (x_min, y_min), (x_max, y_max), color, thickness)

      # Add label text
      label_text = f"{label}"
      font = cv2.FONT_HERSHEY_SIMPLEX
      font_scale = 0.5
      font_thickness = 1
      text_size = cv2.getTextSize(label_text, font, font_scale, font_thickness)[0]
      img_arr = cv2.putText(img_arr, label_text, (x_min, y_min - 5), font, font_scale, color, font_thickness)

  # Display the annotated image
  plt.subplot(1, 2, 2)
  plt.imshow(cv2.cvtColor(img_arr, cv2.COLOR_BGR2RGB))
  plt.title("Annotated Image")
  plt.axis('off')

  plt.show()


# sample 1
image_fp   = data_yolo + "/test/images/image_0.jpg"
label_fp   = data_yolo + "/test/labels/image_0.txt"
visualize_image(image_fp, label_fp)


# sample 2
image_fp   = data_yolo + "/test/images/image_15.jpg"
label_fp   = data_yolo + "/test/labels/image_15.txt"
visualize_image(image_fp, label_fp)


# sample 3
image_fp   = data_yolo + "/test/images/image_105.jpg"
label_fp   = data_yolo + "/test/labels/image_105.txt"
visualize_image(image_fp, label_fp)


## LOAD YOLO V8 NANO BASE MODEL

########################################
#
#   BASE MODEL = "yolov8n.pt"
#
########################################


# Base model file path
base_model_fp = model_fp + "/yolov8n.pt"


# Check if model already exists or not...
if Path(base_model_fp).exists():
  print("`Yolo v8 nano` base model Exists. Loading model...")
  base_model = YOLO(base_model_fp)
  print("\n\n*** Model Loaded! **")
else:
  print("Base model couldn't be found! Please download...")

`Yolo v8 nano` base model Exists. Loading model...


*** Model Loaded! **


# Let's predic
def predict(model, image_fp):

  # Model Inference
  results = model(image_fp)

  # Display
  for r in results:
    #print(r.boxes)
    if len(r.boxes.cls):
      label = int(r.boxes.cls[0])
      print(model.names.get(label))
      im_array = r.plot()  # plot a BGR numpy array of predictions
      im = Image.fromarray(im_array[..., ::-1])  # RGB PIL image
      im.show()
      plt.imshow(im)
    else:
      print("\n\nFailed to detect this image.")


# direct-shot prediction - try #1

predict(base_model, image_fp = data_fp + "/stage2_images/test/images/image_501.jpg")

WARNING ⚠️ NMS time limit 0.550s exceeded
image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_501.jpg: 640x480 1 spoon, 1 dining table, 1 book, 63.4ms
Speed: 7.0ms preprocess, 63.4ms inference, 1977.4ms postprocess per image at shape (1, 3, 640, 480)
spoon


# direct-shot prediction - try #2

predict(base_model, image_fp = data_fp + "/stage2_images/test/images/image_502.jpg")

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_502.jpg: 640x480 3 persons, 9.4ms
Speed: 3.0ms preprocess, 9.4ms inference, 1.8ms postprocess per image at shape (1, 3, 640, 480)
person


# direct-shot prediction - try #3

predict(base_model, image_fp = data_fp + "/stage2_images/test/images/image_504.jpg")

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_504.jpg: 640x480 1 book, 7.2ms
Speed: 2.3ms preprocess, 7.2ms inference, 1.6ms postprocess per image at shape (1, 3, 640, 480)
book


## Fine-tune the Yolo Base Nano Model on our dataset - Playing Cards!


# Load the base model which shall be fine-tuned...
model = YOLO(base_model_fp)
print(f"\n\nBase model loaded:\n{base_model_fp}\n\n\n")

# Training model
print("\n\n<<<<<<<< FINE TUNING... >>>>>>>>>>>>\n\n")
results = model.train(data    = data_fp + "/stage2_images/pokermate.yaml",
                      imgsz   = 416,
                      epochs  = 100,
                      batch   = 250,
                      name    = output_fp + "/pokermate_stage2_yolo8_best.pt")

# Save the model
model.export()


Base model loaded:
/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/models/yolov8n.pt


<<<<<<<< FINE TUNING... >>>>>>>>>>>>


Ultralytics YOLOv8.0.222 🚀 Python-3.10.12 torch-2.1.0+cu118 CUDA:0 (Tesla T4, 15102MiB)
engine/trainer: task=detect, mode=train, model=/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/models/yolov8n.pt, data=/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/pokermate.yaml, epochs=10, patience=50, batch=250, imgsz=416, save=True, save_period=-1, cache=False, device=None, workers=8, project=None, name=pokermate_stage2_yolo8_best.pt, exist_ok=False, pretrained=True, optimizer=auto, verbose=True, seed=0, deterministic=True, single_cls=False, rect=False, cos_lr=False, close_mosaic=10, resume=False, amp=True, fraction=1.0, profile=False, freeze=None, overlap_mask=True, mask_ratio=4, dropout=0.0, val=True, split=val, save_json=False, save_hybrid=False, conf=None, iou=0.7, max_det=300, half=False, dnn=False, plots=True, source=None, vid_stride=1, stream_buffer=False, visualize=False, augment=False, agnostic_nms=False, classes=None, retina_masks=False, show=False, save_frames=False, save_txt=False, save_conf=False, save_crop=False, show_labels=True, show_conf=True, show_boxes=True, line_width=None, format=torchscript, keras=False, optimize=False, int8=False, dynamic=False, simplify=False, opset=None, workspace=4, nms=False, lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=7.5, cls=0.5, dfl=1.5, pose=12.0, kobj=1.0, label_smoothing=0.0, nbs=64, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0, cfg=None, tracker=botsort.yaml, save_dir=/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/output/pokermate_stage2_yolo8_best.pt
Downloading https://ultralytics.com/assets/Arial.ttf to '/root/.config/Ultralytics/Arial.ttf'...

100%|██████████| 755k/755k [00:00<00:00, 93.1MB/s]

Overriding model.yaml nc=80 with nc=52

                   from  n    params  module                                       arguments                     
  0                  -1  1       464  ultralytics.nn.modules.conv.Conv             [3, 16, 3, 2]                 
  1                  -1  1      4672  ultralytics.nn.modules.conv.Conv             [16, 32, 3, 2]                
  2                  -1  1      7360  ultralytics.nn.modules.block.C2f             [32, 32, 1, True]             
  3                  -1  1     18560  ultralytics.nn.modules.conv.Conv             [32, 64, 3, 2]                
  4                  -1  2     49664  ultralytics.nn.modules.block.C2f             [64, 64, 2, True]             
  5                  -1  1     73984  ultralytics.nn.modules.conv.Conv             [64, 128, 3, 2]               
  6                  -1  2    197632  ultralytics.nn.modules.block.C2f             [128, 128, 2, True]           
  7                  -1  1    295424  ultralytics.nn.modules.conv.Conv             [128, 256, 3, 2]              
  8                  -1  1    460288  ultralytics.nn.modules.block.C2f             [256, 256, 1, True]           
  9                  -1  1    164608  ultralytics.nn.modules.block.SPPF            [256, 256, 5]                 
 10                  -1  1         0  torch.nn.modules.upsampling.Upsample         [None, 2, 'nearest']          
 11             [-1, 6]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 12                  -1  1    148224  ultralytics.nn.modules.block.C2f             [384, 128, 1]                 
 13                  -1  1         0  torch.nn.modules.upsampling.Upsample         [None, 2, 'nearest']          
 14             [-1, 4]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 15                  -1  1     37248  ultralytics.nn.modules.block.C2f             [192, 64, 1]                  
 16                  -1  1     36992  ultralytics.nn.modules.conv.Conv             [64, 64, 3, 2]                
 17            [-1, 12]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 18                  -1  1    123648  ultralytics.nn.modules.block.C2f             [192, 128, 1]                 
 19                  -1  1    147712  ultralytics.nn.modules.conv.Conv             [128, 128, 3, 2]              
 20             [-1, 9]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 21                  -1  1    493056  ultralytics.nn.modules.block.C2f             [384, 256, 1]                 
 22        [15, 18, 21]  1    761452  ultralytics.nn.modules.head.Detect           [52, [64, 128, 256]]

Model summary: 225 layers, 3020988 parameters, 3020972 gradients, 8.2 GFLOPs

Transferred 319/355 items from pretrained weights
TensorBoard: Start with 'tensorboard --logdir /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/output/pokermate_stage2_yolo8_best.pt', view at http://localhost:6006/
Freezing layer 'model.22.dfl.conv.weight'
AMP: running Automatic Mixed Precision (AMP) checks with YOLOv8n...
Downloading https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8n.pt to 'yolov8n.pt'...

100%|██████████| 6.23M/6.23M [00:00<00:00, 329MB/s]

WARNING ⚠️ NMS time limit 0.550s exceeded
AMP: checks passed ✅

train: Scanning /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/train/labels.cache... 4000 images, 0 backgrounds, 0 corrupt: 100%|██████████| 4000/4000 [00:00<?, ?it/s]

albumentations: Blur(p=0.01, blur_limit=(3, 7)), MedianBlur(p=0.01, blur_limit=(3, 7)), ToGray(p=0.01), CLAHE(p=0.01, clip_limit=(1, 4.0), tile_grid_size=(8, 8))

val: Scanning /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/val/labels.cache... 700 images, 0 backgrounds, 0 corrupt: 100%|██████████| 700/700 [00:00<?, ?it/s]


## Final model training on 250 epochs...


ImageDisplay(filename=output_fp + '/final_model_1.png', width=800)


ImageDisplay(filename=output_fp + '/final_model_2.png', width=800)


ImageDisplay(filename=output_fp + '/final_model_3.png', width=800)


## Now the model has been fine-tuned!!!

src  = """/content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/output/pokermate_stage2_yolo8.pt/weights/best.pt"""
dest = model_fp + "/pokermate_stage2_yolo8.pt"

# shutil.copy(src, dest)

## check model
if Path(model_fp + "/pokermate_stage2_yolo8.pt").exists():
  print("Fine-tuned model found!")

Fine-tuned model found!


## Let's use the fine-tuned model now!!


# Loading fine-tuned model for cards:
model = YOLO(model_fp + '/pokermate_stage2_yolo8.pt')
print("\n\n** Fine-tuned model loaded! **\n\n")


** Fine-tuned model loaded! **


## original image

img_fp  = data_fp + "/stage2_images/test/images/image_502.jpg"
img_arr = cv2.imread(img_fp)
plt.imshow(img_arr)

<matplotlib.image.AxesImage at 0x79060df49630>


## inferenced image

# prediction
results = model(img_fp)

# View results
for r in results:
    #print(r.boxes)
    label = int(r.boxes.cls[0])
    im_array = r.plot()  # plot a BGR numpy array of predictions
    im = Image.fromarray(im_array[..., ::-1])  # RGB PIL image
    im.show()
    plt.imshow(im)
    im.save(output_fp + '/results.jpg')

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_502.jpg: 416x320 2 9Hs, 12.9ms
Speed: 1.5ms preprocess, 12.9ms inference, 2.7ms postprocess per image at shape (1, 3, 416, 320)


## another!

img_fp  = data_fp + "/stage2_images/test/images/image_503.jpg"
# prediction
results = model(img_fp)
# View results
for r in results:
    #print(r.boxes)
    label = int(r.boxes.cls[0])
    im_array = r.plot()  # plot a BGR numpy array of predictions
    im = Image.fromarray(im_array[..., ::-1])  # RGB PIL image
    im.show()
    plt.imshow(im)
    im.save(output_fp + '/results.jpg')

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_503.jpg: 416x320 2 10Ss, 9.4ms
Speed: 1.3ms preprocess, 9.4ms inference, 2.1ms postprocess per image at shape (1, 3, 416, 320)


# Lets visualize some predictions...

test_data_fp = data_fp + "/stage2_images/test/images/"
testing_data = os.listdir(test_data_fp)

plt.figure(figsize=(16,16))
for i in range(16):
    plt.subplot(4,4,i+1)
    plt.xticks([])
    plt.yticks([])
    plt.grid(False)

    # pick a random image
    index = np.random.randint(0, len(testing_data))
    img_name = testing_data[index]
    print(img_name)
    img_fp = test_data_fp + img_name
    #plt.imshow(img_fp)

    # inference
    results = model(img_fp)

    # View results
    for r in results:
        #print(r.boxes)
        try:
          # GPU VERSION
          labels = set(np.array(r.boxes.cls).astype(int))
        except Exception as e:
          # CPU VERSION
          labels = set(np.array(r.boxes.cls.cpu()).astype(int))
        class_labels = [model.names.get(x) for x in labels]
        im_array = r.plot()
        im = Image.fromarray(im_array[..., ::-1])
        im.show()
        plt.imshow(im)
        plt.xlabel(f"Labels: {class_labels}")

plt.show()

image_83.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_83.jpg: 416x416 4 2Hs, 1 4H, 1 7S, 15.7ms
Speed: 1.1ms preprocess, 15.7ms inference, 3.1ms postprocess per image at shape (1, 3, 416, 416)
image_340.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_340.jpg: 416x416 4 2Ds, 3 5Ss, 11.0ms
Speed: 1.0ms preprocess, 11.0ms inference, 2.4ms postprocess per image at shape (1, 3, 416, 416)
image_111.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_111.jpg: 416x416 1 3H, 4 3Ss, 1 9H, 15.4ms
Speed: 1.0ms preprocess, 15.4ms inference, 2.7ms postprocess per image at shape (1, 3, 416, 416)
image_1.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_1.jpg: 416x416 1 6S, 3 JSs, 1 KC, 11.0ms
Speed: 1.2ms preprocess, 11.0ms inference, 2.6ms postprocess per image at shape (1, 3, 416, 416)
image_491.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_491.jpg: 416x416 4 3Ss, 4 5Cs, 16.7ms
Speed: 1.1ms preprocess, 16.7ms inference, 2.9ms postprocess per image at shape (1, 3, 416, 416)
image_443.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_443.jpg: 416x416 3 10Cs, 4 5Cs, 34.5ms
Speed: 2.1ms preprocess, 34.5ms inference, 3.1ms postprocess per image at shape (1, 3, 416, 416)
image_121.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_121.jpg: 416x416 1 8C, 1 9D, 4 KHs, 13.6ms
Speed: 1.3ms preprocess, 13.6ms inference, 2.6ms postprocess per image at shape (1, 3, 416, 416)
image_265.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_265.jpg: 416x416 1 7D, 1 9S, 4 AHs, 10.7ms
Speed: 0.9ms preprocess, 10.7ms inference, 2.0ms postprocess per image at shape (1, 3, 416, 416)
image_83.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_83.jpg: 416x416 4 2Hs, 1 4H, 1 7S, 10.7ms
Speed: 0.9ms preprocess, 10.7ms inference, 2.0ms postprocess per image at shape (1, 3, 416, 416)
image_70.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_70.jpg: 416x416 4 10Ss, 3 4Hs, 10.8ms
Speed: 0.9ms preprocess, 10.8ms inference, 2.2ms postprocess per image at shape (1, 3, 416, 416)
image_279.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_279.jpg: 416x416 3 6Ds, 4 6Ss, 10.9ms
Speed: 1.0ms preprocess, 10.9ms inference, 2.2ms postprocess per image at shape (1, 3, 416, 416)
image_247.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_247.jpg: 416x416 1 2D, 3 JHs, 1 JS, 10.7ms
Speed: 0.9ms preprocess, 10.7ms inference, 2.0ms postprocess per image at shape (1, 3, 416, 416)
image_210.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_210.jpg: 416x416 1 10S, 4 5Hs, 1 JH, 10.6ms
Speed: 0.8ms preprocess, 10.6ms inference, 2.4ms postprocess per image at shape (1, 3, 416, 416)
image_352.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_352.jpg: 416x416 2 KHs, 2 QDs, 2 QSs, 10.6ms
Speed: 0.9ms preprocess, 10.6ms inference, 1.9ms postprocess per image at shape (1, 3, 416, 416)
image_171.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_171.jpg: 416x416 1 9C, 1 9S, 2 KHs, 10.7ms
Speed: 0.9ms preprocess, 10.7ms inference, 2.0ms postprocess per image at shape (1, 3, 416, 416)
image_73.jpg

image 1/1 /content/drive/MyDrive/WORK/Data Science/7.  Neural Networks and Deep Learning/resources/CSCI 5922 - Neural Nets and Deep Learning/ASSIGNMENTS/PROJECT/source/ml/data/stage2_images/test/images/image_73.jpg: 416x416 1 6D, 2 JCs, 1 QC, 10.8ms
Speed: 0.9ms preprocess, 10.8ms inference, 2.1ms postprocess per image at shape (1, 3, 416, 416)


# final model architecture:

model

YOLO(
  (model): DetectionModel(
    (model): Sequential(
      (0): Conv(
        (conv): Conv2d(3, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (1): Conv(
        (conv): Conv2d(32, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (2): C2f(
        (cv1): Conv(
          (conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(96, 64, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (3): Conv(
        (conv): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (4): C2f(
        (cv1): Conv(
          (conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0-1): 2 x Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (5): Conv(
        (conv): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (6): C2f(
        (cv1): Conv(
          (conv): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0-1): 2 x Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (7): Conv(
        (conv): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (8): C2f(
        (cv1): Conv(
          (conv): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(768, 512, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (9): SPPF(
        (cv1): Conv(
          (conv): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): MaxPool2d(kernel_size=5, stride=1, padding=2, dilation=1, ceil_mode=False)
      )
      (10): Upsample(scale_factor=2.0, mode='nearest')
      (11): Concat()
      (12): C2f(
        (cv1): Conv(
          (conv): Conv2d(768, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(384, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (13): Upsample(scale_factor=2.0, mode='nearest')
      (14): Concat()
      (15): C2f(
        (cv1): Conv(
          (conv): Conv2d(384, 128, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(192, 128, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (16): Conv(
        (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (17): Concat()
      (18): C2f(
        (cv1): Conv(
          (conv): Conv2d(384, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(384, 256, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (19): Conv(
        (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
        (act): SiLU(inplace=True)
      )
      (20): Concat()
      (21): C2f(
        (cv1): Conv(
          (conv): Conv2d(768, 512, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (cv2): Conv(
          (conv): Conv2d(768, 512, kernel_size=(1, 1), stride=(1, 1))
          (act): SiLU(inplace=True)
        )
        (m): ModuleList(
          (0): Bottleneck(
            (cv1): Conv(
              (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (cv2): Conv(
              (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
          )
        )
      )
      (22): Detect(
        (cv2): ModuleList(
          (0): Sequential(
            (0): Conv(
              (conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
          )
          (1): Sequential(
            (0): Conv(
              (conv): Conv2d(256, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
          )
          (2): Sequential(
            (0): Conv(
              (conv): Conv2d(512, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
          )
        )
        (cv3): ModuleList(
          (0): Sequential(
            (0): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(128, 52, kernel_size=(1, 1), stride=(1, 1))
          )
          (1): Sequential(
            (0): Conv(
              (conv): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(128, 52, kernel_size=(1, 1), stride=(1, 1))
          )
          (2): Sequential(
            (0): Conv(
              (conv): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (1): Conv(
              (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
              (act): SiLU(inplace=True)
            )
            (2): Conv2d(128, 52, kernel_size=(1, 1), stride=(1, 1))
          )
        )
        (dfl): DFL(
          (conv): Conv2d(16, 1, kernel_size=(1, 1), stride=(1, 1), bias=False)
        )
      )
    )
  )
)

PokerMate - Stage 2: Object Detection¶

Problem Statement¶

Approach¶

Dataset¶

1. Setup¶

2. Imports¶

3. Dataset¶

4. Visualization¶

5. Transfer Learning using YOLO (Version 8)¶

5.1. Direct-shot Prediction (No fine-tuning)¶

6. Fine-tuning¶

7. Model Prediction¶